2026 ASEE Annual Conference & Exposition

Rapid advances in artificial intelligence (AI) call for engineering curricula that foster systems thinking and practical problem-solving skills in an interdisciplinary environment, areas often underserved by coding-centric computer science courses. This paper presents a systems-based approach to AI education that frames AI not as a collection of isolated algorithms, but as an interconnected lifecycle grounded in stakeholder values and real-world constraints. The approach has been implemented in a semester-long undergraduate course (AI for Smart Cities) and a week-long summer program. Course is structured around three mutually reinforcing elements: (1) AI lifecycle modules, which include Identify User Needs, Formulate AI Tasks, Select and Develop Models, Manage Data, Evaluate Performance, Analyze Errors, Refine with Feedback, Complete Solution, and Feasibility Analysis; (2) explicit inter-module relationships, such as dependencies, iteration, and error propagation across the system; and (3) critical information artifacts within each module, created using critical-thinking methods with characterization, categorization, and prioritization. Value-Sensitive Design principles (e.g., accessibility, public safety, resilience) are integrated throughout the lifecycle through stakeholder maps and trade-off matrices, ensuring that design choices are justified based on stakeholder needs. Team-based labs, peer review, and real-world projects support interdisciplinary collaboration. Preliminary self-reported pre/post survey results indicated perceived gains in students’ understanding of the systems approach and related AI problem-solving competencies. The systems-based approach simplifies complex theories into accessible concepts, helping students build competencies in identifying key modules, reasoning about inter-module relationships, and critically evaluating information.